Like so many other things, context matters. In some cases isolates and synthetics are not as effective or safe. Some substances are either not as effective or not active at all without an adjunct/adjudivant. However, pretty much any common medication is indeed safer as an isolate/synthetic.

A potential counter example would be willow bark for an analgesic. The plant doesn’t contain acetylsalicylic acid but the prodrug salicin. However, the concentration of salicin typically found in willow is too low to be of significant use. It still works despite this though because of the flavonoids and polyphenols in the bark augment the analgesic and anti-inflammatory action of the salicylates produced by salicin metabolism. This actually provides a broader range of action than aspirin, while having a much better side effect profile, notably lacking risk to the digestive tract.

Then, you can look at foxglove (aka digitalis). I had a neighbor growing up that suffered from cardiac arrest caused by digitalis poisoning from handling (and likely eating) foxglove as a child. He lived but had permanent brain damage. However, digoxin extracted and purified from foxglove is incredibly useful for treating atrial fibrillation in CHF patients.

Where it gets really cool to me though is taking a step back and looking at biological/microbiological manufacturing processes and the impact of “superfluous” chemicals at that point. For example, pu ehr tea goes through pretty complex fermentation and genetic tests have found species of fungi that produce mycotoxins of concern. However, in the presence of chemicals found in tea leaves, production of said toxins is suppressed below the threshold of detection. This results in drinkable tea with potentially beneficial chemicals from plant, fungal, and bacterial origins.